Drupal 和 WordPress 开发日志:贡献信任、浏览器 KEV、GPL 洗涤与 1M-Context 代码审查

📄 中文摘要

开放会员制度在主要风险为粗糙合并时有效,但当维护者需要在大规模过滤机器生成的垃圾时,这一制度迅速崩溃。这一问题不仅限于 Python,Drupal 的贡献和 WordPress 插件生态系统同样依赖于脆弱的人类注意力。AI 垃圾邮件已成为贡献维护者面临的治理问题,GitHub 的“混乱大潮”使得 Jazzband 的开放会员和共享推送访问模型变得不可行。

📄 English Summary

Drupal and WordPress Devlog: Contrib Trust, Browser KEVs, GPL Washing, and 1M-Context Code Review

The open membership model was effective when the main risk was a sloppy merge, but it quickly breaks down when maintainers are tasked with filtering machine-generated junk at scale. This issue extends beyond Python, as both the Drupal contrib and WordPress plugin ecosystems rely on fragile human attention. AI spam has now become a governance problem for contrib maintainers, with GitHub's 'slopocalypse' rendering Jazzband's model of open membership and shared push access untenable.

Powered by Cloudflare Workers + Payload CMS + Claude 3.5

数据源: OpenAI, Google AI, DeepMind, AWS ML Blog, HuggingFace 等